13. Meta-analysis of clinical trials

  1. A. Gouveia Oliveira

Published Online: 12 JUL 2013

DOI: 10.1002/9781118670767.ch13

Biostatistics Decoded

Biostatistics Decoded

How to Cite

Oliveira, A. G. (ed) (2013) Meta-analysis of clinical trials, in Biostatistics Decoded, John Wiley & Sons, Ltd, Chichester, UK. doi: 10.1002/9781118670767.ch13

Editor Information

  1. Department of Pharmacy, Universidade Federal do Rio Grande do Norte, Natal, RN, Brazil

Publication History

  1. Published Online: 12 JUL 2013
  2. Published Print: 3 SEP 2013

ISBN Information

Print ISBN: 9781119953371

Online ISBN: 9781118670767

SEARCH

Keywords:

  • clinical trials;
  • heterogeneity;
  • homogeneity;
  • inverse variance method;
  • meta-analysis;
  • publication bias;
  • random effects model;
  • treatment effect

Summary

The most common application of meta-analysis of clinical trials is the combination of results of inconclusive clinical trials with the purpose of increasing the sample size and thereby conferring increased power to the statistical testing of treatment effects. Meta-analysis is based on a weighted average of the treatment effects. Inverse variance method gives us a weighted estimate of the true treatment effect, its confidence interval, and a test of homogeneity of treatment effects across trials. The random effects model assumes that each clinical trial represents a measurement of the true treatment difference and that the results obtained in the several clinical trials differ only because of sampling variation. The test of heterogeneity has low statistical power and a non-significant result by no means excludes the possibility of heterogeneity. The important reason for publication bias is the greater likelihood of rejection of negative studies by editors of scientific journals.